@Borrecan This code replaces SoundAnalysis with AudioSession and AudioEngine, implementing the entire audio capture and analysis workflow from scratch. Audio is captured through the AudioEngine’s inputNode, then processed using FFT (Fast Fourier Transform) to convert the signal from the time domain to the frequency domain.

@Borrecan Features within the target frequency range are extracted and packed into an MLMultiArray, which is passed to a custom CoreML model (BeepRecogniserIOS) for classification to determine whether the sound matches the target (e.g., a beep).

@anthonyshook Compared to SoundAnalysis, this approach is more flexible, allowing custom audio processing logic, frequency range adjustments, and the use of a tailored model for more accurate detection. While slightly more complex to implement, it offers greater control and is especially suitable for scenarios requiring specialized audio detection.